1 / 25

Why Cryptosystems Fail

Why Cryptosystems Fail. Nick Feamster CS 6262 Spring 2009. Overview. Case Study: Banking Industry and ATM Fraud Fraud Types Insider Knowledge and Access Outside Attacks with Minimal Technical Knowledge Banks without Security Modules Inferior Security Products

levi
Télécharger la présentation

Why Cryptosystems Fail

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Why Cryptosystems Fail Nick FeamsterCS 6262Spring 2009

  2. Overview • Case Study: Banking Industry and ATM Fraud • Fraud Types • Insider Knowledge and Access • Outside Attacks with Minimal Technical Knowledge • Banks without Security Modules • Inferior Security Products • Poor Implementation and/or Sloppy Operating Procedures

  3. Overview • Consequences for Bankers • Implications for Equipment vendors • Why the Threat Model was Wrong? • Authors Confirmation of this • Shift to New Security Paradigm? • Conclusions/Discussion

  4. Cryptology Intro • Main User of Cryptosystems • Government Agencies • NSA • Generally do not release failure reports • Puts Cryptographic Designers at a severe disadvantage • Removes Learning Mechanism

  5. Cryptology Intro • Since Government Agencies are very secretive about their mistakes • We must look elsewhere for examples • Next largest application of Cryptology • Retail Banking Systems • Therefore we’ll take a look their failure modes

  6. Case Study: ATM Fraud • Inside Knowledge or Access • Bank Stat • Bank teller/clerk issue extra card • Maintenance Engineer retrofitted an ATM with a mini PC to record account numbers and PIN’s • Then produced counterfeit cards • Dual Control of cards and pins dropped to cut costs

  7. Case Study: ATM Fraud • Outsider Attacks • Simple yet Effective: Observing customers entering PIN’s and pick up discarded receipts • Another Example – telephone card • Jackpotting • ATM networks do not encrypt/authenticate the authorization response to the ATM • Attacker can then record a ‘pay’ response and replay it until the ATM is empty

  8. Case Study: ATM Fraud • Outsider Attacks • Postal Interception • Similar to Credit Cards • Test Transactions in ATM’s • Implementation error, sequence printed in branch manual • False Terminals • Harvest account numbers and PIN’s from unknowing customers

  9. Case Study: ATM Fraud • Outsider Attacks • PIN Security Reductions • PIN’s checked and set by offline devices • Devices do not always use full encryption • Insecure bank suggestions • Random letter card • Reduced chances of guessing PIN from 1 in 3,333 to 1 in 8 • Using default PIN numbers • Result of a simple programming error

  10. Case Study: ATM Fraud • Outsider Attacks • PIN’s not derived from account number • Encrypted PIN’s on a file • Programmer could look for other accounts with the same encrypted PIN as their own • Encrypted PIN’s written to the card stripe • Thief could then change the account number of their card and access and other account using their PIN

  11. Case Study: ATM Fraud • How ATM Encryption Works • Account Number: 997883242329 • PIN Key: GGFFGGFFFGGF • Result of DES: R2FF3Y2430TG • Result decimalized: 07773248233 • Natural PIN: 0777 • Offset: 1111 • Customer PIN: 1888

  12. Case Study: ATM Fraud • Problems with encryption products • Banks without Security Modules • Encryption is then handled in SW • Biggest problem is that the PIN key can easily be found by system programmers • Even if security is added at a later date is unlikely to fix the problem • PIN key is so valuable to networked ATM’s that knowledge will likely remain among programming staff

  13. Case Study: ATM Fraud • Inferior Security Products • Backdoors in security module’s SW • Security modules enclosures could be comprised • Often possible to penetrated by drilling or cutting • Tamper protection implemented with wires leading to switches • Maintenance engineer could easily cut, then have access to keys on next visit

  14. Case Study: ATM Fraud • Poor Implementation or Operating Procedures • Ignoring response codes • ‘Key Parity Error’ programmer altering live module • Giving the PIN key to a facilities management firm • Employee turnover • Outside firms may not share banks security culture

  15. Case Study: ATM Fraud • Poor Implementation or Operating Procedures • Poor Design Psychology • Banks end up cutting corners and share sensitive information • “Security by Obscurity” often does more harm than good • Not properly incorporated in operational procedures • Should be explicitly stated in all manuals and training courses

  16. Case Study: ATM Fraud • Cryptanalysis • Banks using weak Encryption Algorithms • Home-grown algorithms • Pre-DES algorithms • Respectable Algorithm poorly implemented • Using a key that is too small to provide the necessary security • Writing the PIN to the card track • DES Keys can be found by brute force • Once found all PINs could be decrypted • Countries in chaos have proper equipment

  17. Consequences for Bankers • Original goal of ATM crypto security • No system fraud could take place without at least two bank staff • Why has this not happened? • Poor Implementation • Unorganized and Uniformed Administration • Higher emphasis on Quality Control

  18. Consequences for Bankers • Concerns for the Future • Current Security Systems • Built from components that were not well understood • Admin support requirements not clearly defined • These current issues will slow transition • Making it harder to integrate new components • Decreasing the likelihood of proper long term maintenance

  19. Implications for Equipment Vendors • Re-evaluate the system level approach to designing and evaluating security • What should they design towards? • Autonomous Systems • Modifiable Systems that require proper administration and maintenance • Develop a certification process to address the human environment that the system will operate

  20. Implications for Equipment Vendors • Three courses of action • Designs systems that can be easily integrated and maintained by a general computer staff • Train and certify client personnel to perform integration and maintenance • Provide their own personnel to implement, support and manage the system

  21. Why the Threat Model was Wrong? • Threat was misjudged • Expected criminals with a high level of technical expertise • Customers’ Abilities severely Misjudged • Assumed the implementers at customer sites would have appropriate expertise for the job • Or that they would enlist the help of quality consulting to complete the task

  22. Why the Threat Model was Wrong? • Why were the Threat and Customer Abilities so badly misjudged? • Companies imported help from the military sector • The military model strictly focused on security • Human factors • Many organizations security teams were nonexistent or limited • Some Consulting firms miss represent their technical expertise in the area of computer security

  23. Confirmation of this Analysis • The military sector has had the exact same experiences: • Specifically reported by an senior NSA scientist • The vast majority of failures are occurring at the level of implementation • The NSA is not more clever than the civilian population, just better informed • The threat profiles developed by the NSA for its own use are classified

  24. Shift to New Security Paradigm? • Evidence that a shift is necessary • Shift from building evaluated products to focusing on quality control products within the client • Organization • Author suggests a metaphor to safety critical systems • Competing Philosophies • Railway Signaling Systems vs. Aviation Paradigm

  25. Conclusions / Discussion • Why Have Cryptosystems been Failing? • Cryptosystems designers have a lack of feedback on how there systems fail • Therefore they’ve been designing towards the wrong end • Many security products are so complex and tricky to use the are rarely used properly • As a result most security failures are due to implementation and management errors

More Related